|
|
||||||||

T Cell Subset1,2


* Department of Medicine, The Johns Hopkins University School of Medicine, Baltimore, MD 21205;
Unité dImmunité Cellulaire Antivirale Département SIDA-Rétrovirus, Institut Pasteur, Paris, France; and
Department of Microbiology and Immunology, F. Edward Hebert School of Medicine, Uniformed Services University of the Health Sciences, Bethesda, MD 20814
| Abstract |
|---|
|
|
|---|

+CD94TCR
+ T cell subset within the iIEL population in response to oral infection with virulent or avirulent Salmonella. This population can be detected 3 days following infection, represents up to 15% of the TCR
+ iIELs, and is dependent on the MHC class Ib molecule T23 (Qa-1). Qa-1 is expressed by intestinal epithelial cells and thus accessible for iIEL recognition. Such cells may play a role in the early immune response to Salmonella. | Introduction |
|---|
|
|
|---|
T cells of the adaptive immune system play a central role in the clearance of primary infections and in protection against subsequent challenge with related strains of Salmonella (7, 8). T cell-deficient mice are impaired in their ability to clear primary oral infections with virulent Salmonella dublin (9), and athymic mice fail to clear systemic primary infections caused by various attenuated S. typhimurium strains (10, 11). Salmonella reside intracellularly within phagocytic and nonphagocytic cells, and bacterial proteins can gain access to both the MHC class I and MHC class II Ag-processing pathways (12, 13, 14, 15). Thus, both CD4+ and CD8+ T cells have been implicated in the immune response to Salmonella (16, 17, 18, 19, 20, 21).
T cells that reside in the intestinal epithelium (intestinal intraepithelial lymphocytes (iIELs)) are the first elements of the host T cell compartment available to respond to Salmonella during its entry into the host by oral infection. iIELs can be divided into two main subpopulations based on CD8 expression. One subpopulation expresses the TCR
and CD8
(22, 23). The features of the TCR repertoire expressed by these T cells suggest that they have undergone thymic selection, and they are absent in athymic mice (24, 25, 26). The CD8
TCR
cells within the iIEL population and in peripheral lymphoid organs are largely dependent on MHC class Ia molecules and TAP for their development (27, 28, 29).
The other subpopulation of T cells found within the iIEL compartment expresses TCR
or TCR
in combination with CD8
. CD8
iIELs are present in athymic mice, and their development is largely confined to the intestine (25), potentially occurring in gut-associated cryptopatches (30). Intestinal TCR
CD8
T cell development is MHC class I dependent, and partially TAP dependent (31, 32, 33). However, although class I dependent, this iIEL population is MHC class Ia and CD1 independent (27, 34, 35), implying that MHC class Ib molecules, such as Qa-2, may play a role in their development (36).
TCR
+ cells within the iIEL population (
iIELs) predominantly express CD8
, and a minor population does not express CD4 or CD8 (37). The development of most 
iIELs is independent of MHC class I molecules (38). However, some MHC class Ib molecules are recognized by TCR
+ cells within the iIEL population, either via their TCR or through other cell surface receptors. For example, a subset of 
iIELs recognizes the closely related molecules T10 and T22 via their TCRs (39, 40, 41), and the MHC class Ib molecule TL interacts with 
iIELs through binding to CD8
(42). Subsets of TCR
+ cells have also been shown to express NKG2D, a receptor for the MHC class I-related RAE-1 and H-60 proteins (43). Murine CD1d, T22, and TL are expressed by intestinal epithelial cells (44, 45, 46), providing further evidence that MHC class Ib molecules may be able to directly modulate the effector function of intestinal intraepithelial TCR
+ cells.
In this study, we report the Qa-1-dependent expansion of a novel CD8
+TCR
+ cell subset within the iIEL population, during the initial stages of the immune response to oral infection with virulent or avirulent S. typhimurium strains. Qa-1 is shown to be expressed by intestinal epithelial cells; thus, this CD8
+TCR
+ cell subset, and hence Qa-1, may play a role in the mucosal intestinal immune response to Salmonella.
| Materials and Methods |
|---|
|
|
|---|
C57BL/6 (B6, wild-type) mice were purchased from either The Jackson Laboratory (Bar Harbor, ME) or the National Cancer Institute (Frederick, MD). B6-
2m/ and B6-Tap-1/ (29, 47) mice were purchased from The Jackson Laboratory. Kb/Db/ were backcrossed onto a B6/J mice, and their generation was described previously (48, 49). Handling of mice and experiments were conducted in accordance with institutional policies for animal care and use.
Bacterial strains and growth conditions
Virulent S. typhimurium strain C5 (wild-type) and live, avirulent S. typhimurium strain SL3235 aro A (aroA) were grown in Bacto tryptic soy (TS) broth (BD Biosciences, Sparks, MD) or on TS agar plates. Bacteria were grown overnight at 37°C. The following day, bacterial cultures at saturation density were diluted 1/100 and grown to mid-logarithmic phase (A600, 0.50.6). Bacterial cultures were centrifuged and washed in PBS twice before use. The number of bacteria was calculated from a standard curve relating CFUs vs OD600 (A600), and verified by growing aliquots of serial dilutions on TS agar plates.
In vivo infection with S. typhimurium
Eight- to 10-wk-old gender-matched mice were either mock infected or infected orally with 5 x 105 wild-type or 5 x 108 aroA Salmonella in 0.1 ml of PBS, using a standard gastric intubation needle. At various times after infection, small intestine, mesenteric lymph nodes (MLN), PP, liver, and spleen were isolated from infected and control animals. The number of bacteria present in each mouse strain was determined by preparing single-cell suspensions from the MLN, PP, liver, and spleen using a Stomacher (Seward, London, U.K.), and making serial dilutions in PBS of the suspensions obtained. Aliquots of each serial dilution were grown on TS agar plates.
Antibodies
Purified anti-CD16/CD32 (clone 2.4G2), FITC and PE-anti-CD8
.2 (clone 53-5.8), PE-anti-TCR
(clone GL3), PerCP anti-CD8
(clone 53-6.7), allophycocyanin-anti-TCR
(clone H57-597), biotinylated anti-CD94 (clone 18d3), biotinylated anti-Qa-1b (clone 6A8), FITC-anti-NK1.1 (clone PK136), biotinylated mouse IgG1, and streptavidin-allophycocyanin were purchased from BD PharMingen (San Diego, CA). Anti-Qa-1b (4C2) was prepared from ascites by protein A/G affinity chromatography using standard protocols. Murine IgG1, FITC-F(ab')2 sheep anti-mouse IgG1, and HRP-F(ab')2 sheep anti-mouse IgG1 were purchased from Accurate Scientific (Westbury, NY). HRP-streptavidin was purchased from DAKO (Carpinteria, CA)
Cell preparation and flow cytometry
iIEL were isolated from the small intestine as described previously (50). In brief, PP were identified and removed. After flushing with PBS (4°C), the gut was opened on a wet linen square. The mucosa was scraped with a scalpel, and then dissociated by stirring in 50 ml of RPMI 1640 containing 10% heat-inactivated FCS, 1 mM dithioerythritol (Sigma-Aldrich, St. Louis, MO), and 1 mM EDTA for 20 min at room temperature. After centrifugation, the pellet was resuspended in RPMI 1640 containing 10% heat-inactivated FCS and passed through a glass wool column prewashed with RPMI 1640 containing 5 mM HEPES. The glass wool-nonadherent cells were centrifuged and resuspended in 72% Percoll (Amersham Pharmacia Biotech, Piscataway, NJ)/PBS solution, overlaid with an equal volume of 36% Percoll/PBS, and centrifuged at 2000 x g for 30 min at room temperature. The cells (iIELs) at the interface were isolated, washed, and counted.
The recovered iIEL population was first incubated with anti-CD16/CD32 to block nonspecific mAb binding. All iIEL and cell line samples were stained with various combinations of directly and indirectly fluorochrome-conjugated mAbs, as indicated. Three- or four-color immunofluorescence staining was analyzed by either FACScan or the FACSCalibur instrument (BD Biosciences). iIELs were gated using forward and side scatter to exclude dead cells, and the data were analyzed using CellQuest software (BD Biosciences).
Immunoblotting
Whole-cell lysates from cell lines or homogenized tissues were prepared by the addition of SDS sample buffer without 2-ME, followed by boiling and sonication. Lipids were extracted from brain and small intestine lysates using 1,1,2-trichlorotrifluoroethane. Protein concentrations were determined using the Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA). Fifteen micrograms of each lysate was reduced by the addition of 2-ME, boiled, separated by SDS-PAGE (10%), and transferred to Immun-Blot polyvinylidene difluoride membrane (Bio-Rad). The presence of Qa-1 was visualized using the mAb 4C2 (anti-Qa-1b), followed by HRP-F(ab')2 sheep anti-mouse IgG1 and ECL (Amersham, Chicago, IL).
Chromium release assay
Lysis of 51Cr-labeled target cells was performed as described (51). In brief, target cells were prepared from the appropriate mouse strains by culturing spleen cells in the presence of 1 µg/ml Con A for 3 days, and live cells were then purified using Lympholyte M (Accurate Scientific). All target cell types were labeled for 1.5 h at 37°C with 51Cr (Amersham Pharmacia Biotech). A total of 1 x 104 target cells per well were aliquoted into 96-well plates (Falcon, Bedford, MA) and pulsed with and without the peptide AMAPRTLLL (Qdm) (Macromolecular Resources, Fort Collins, CO) followed by incubation at 37°C for 1 h. CTL effector cells were added at the indicated E:T ratios for 4 h at 37°C before supernatants were collected and counted on a Microbeta instrument (model 1450; Wallac, Gaithersburg, MD). In all cases, the percent specific lysis at an indicated E:T ratio represents the mean of triplicate samples.
Single-chain Qa-1b (scQa-1b) transgenes
The strategy to generate a single-chain class I molecules has been described in detail previously (52, 53, 54). In brief, scQa-1b transgenes were constructed by linking the C terminus of murine
2mb cDNA to the N terminus of the
2 domain of Qa-1b cDNA (scQa-1b(c)) or Qa-1b genomic DNA (scQa-1b(g)), via a spacer of 15 aa containing the sequence (G4S)3. PCR was used to construct a modified
2mb cDNA sequence, with an upstream primer that contained a sequence corresponding to the proximal end of
2mb, and a downstream primer that contained a sequence corresponding to the distal end of
2mb, followed by a sequence encoding the first 13 aa of the (G4S)3 linker. A modified version of Qa-1b cDNA was constructed using an upstream primer that encoded the last 13 aa of the (G4S)3 linker, followed by a sequence corresponding to the proximal end of exon 2 of Qa-1b, and a downstream primer that contained a sequence corresponding to the 3' untranslated region of Qa-1b. ScQa-1b(c) was generated by mixing the two PCR products and splicing by overlap extension, using the upstream primer specific for the proximal end of
2mb and the downstream primer specific for the 3' untranslated region of Qa-1b. ScQa-1b(c) was then cloned into pCI-neo (Promega, Madison, WI). A fragment of the H-2Kb promoter, from 263 to + 20, was PCR amplified and subcloned into pCI-neo, replacing the existing promoter but leaving the 5' untranslated region chimeric intron intact.
ScQa-1b(g) was created by subcloning a fragment of genomic Qa-1b from the BstXI site, corresponding to a sequence in exon 2, through to the SpeI site, corresponding to a sequence in the 3' untranslated region, into scQa-1b(c). A fragment of the H-2Dd promoter, from XbaI to BamHI,
400 bp upstream of the initiating ATG, was PCR amplified and subcloned directly upstream of the proximal end of scQa-1b(g). Transgene DNA was injected into (B6 x SJL)F1 blastocysts. Founder mice were crossed to B6-
2m/ (47) mice for at least six generations.
Cell lines and transfectants
DLD-1 (American Type Culture Collection, Manassas, VA), a
2-microglobulin (
2m)/ human colorectal tumor cell line (55, 56), was cultured in RPMI 1640 supplemented with 10% heat-inactivated FCS, 1 mM sodium pyruvate, 10 mM HEPES, 100 U/ml penicillin, and 100 µg/ml streptomycin. KJ29, a
2m/ human liver tumor cell line, was maintained in DMEM with the same supplements (57). DLD-1 and KJ29 were transfected with scQa-1b(c) and scQa-1b(g) using Lipofectamine Plus (Invitrogen, Grand Island, NY), and transfectants were selected using medium supplemented with Geneticin (G418) at 300 and 600 µg/ml, respectively.
The human lymphoblastoid cell line (C1R) was maintained in RPMI 1640 supplemented with 5% heat-inactivated bovine calf serum, 10 mM HEPES, 100 U/ml penicillin, and 100 µg/ml streptomycin. C1R cells expressing Qa-1b were generated previously (51). C1R cells expressing scQa-1b(c) and scQa-1b(g) were generated by electroporating 10 µg of linearized versions of both plasmids at 230 V and 960 µF, following a standard protocol, and transfected cells were selected using medium supplemented with 800 µg/ml G418.
The Qa-1b-restricted CTL clone D5D2 was described previously (51). D5D2 recognizes the dominant Qa-1-associated peptide Qdm, which is derived from aa 311 of most MHC class 1a D region-encoded molecules. D5D2 also recognizes other peptides that are very similar in sequence to Qdm, including the Dk variant (AMVPRTLLL), when presented by Qa-1.
Immunohistochemistry
Tissue samples from 8- to 10-wk-old mice were fixed in alcoholic zinc formalin (50% isopropyl alcohol, 8 mM zinc chloride, and 3.7% formalin) and embedded in paraffin. Four-micrometer sections were deparaffinized and rehydrated in graded alcohols before immunohistochemical studies. Biotinylated 6A8 (anti-Qa-1b) or biotinylated mouse IgG1 were overlaid at a final concentration of 25 µg/ml in 50% normal goat serum/50% pooled human serum for 12 h at 4°C. Slides were washed and incubated with HRP-streptavidin at a final concentration of 5 µg/ml for 1 h at room temperature. Diaminobenzidine (DAKO) was used as the chromogen, and hematoxylin (DAKO) was used as the nuclear counterstain. Stained samples were examined using a confocal microscope system (LSM 410; Carl Zeiss, Thornwood, NY). Images were collected and analyzed with the manufacturers software.
| Results |
|---|
|
|
|---|
Previous studies, using an i.p. infection route, demonstrated that class I-deficient (
2m/) mice were more susceptible to infection with S. typhimurium (21). Given that the natural route of infection with Salmonella is through the ingestion of contaminated food, we examined this issue in an oral infection model. As shown in Fig. 1, viable bacteria can be detected in the spleens and livers of wild-type B6 and B6-
2m/ mice following oral infection with wild-type S. typhimurium. In C57BL mice, bacteria can be detected as early as 3 days postinfection, and by 10 days, the in vivo bacterial load has plateaued. In class I-deficient (
2m/) mice, bacterial loads were consistently higher, and bacterial growth continued to rise. Similar observations were made when bacterial loads in MLN and PP were measured (Fig. 1).
|

+TCR
+ iIEL subset
The observation that class I-deficient mice show increased bacterial growth as early as 3 days post-oral infection suggested that class I-dependent elements of the mucosal lymphoid compartment played a role in the ability of the host to limit infection. iIELs are uniquely positioned to respond to infections initiated via the gastrointestinal tract. Therefore, we have analyzed the distribution and/or frequency of iIEL subsets following oral infection with S. typhimurium. Displayed in Fig. 2A is a flow-cytometric analysis of the distribution of TCR
and TCR
iIELs isolated from infected and mock-infected B6 mice. In mock-infected B6 mice, TCR
-expressing cells were the dominant iIEL population. However, B6 mice orally infected with wild-type Salmonella consistently displayed an increase in proportion of TCR
+ iIELs. The yield of iIELs was typically 24 x 106 cell/mouse and did not significantly differ in infected or uninfected mice. Therefore, there was no dramatic infection-induced change in the total number of iIELs recovered.
|

+ iIELs represented a change in a specific subpopulation, iIELs from wild-type mice were analyzed at various time points after oral infection with wild-type Salmonella, and the expression of CD8
and CD8
on TCR
iIELs was examined. As reported previously (25, 38), in uninfected mice, the TCR
-expressing iIELs largely expressed CD8
(Fig. 2B). However, in iIELs recovered from infected mice, a TCR
+ cell population expressing CD8
and CD8
is readily detected. This CD8
+TCR
+ iIEL population is variably seen in uninfected animals, yet by 10 days postinfection represented up to 15% of the total TCR
+ iIEL population. Examination of TCR
+ cells from PP, MLN, livers, or spleens failed to identify a CD8
+-expressing TCR
+ subset. Thus, this subpopulation was uniquely found in the intestinal epithelial compartment.
To determine whether the expansion of CD8
+TCR
+ iIEL population is affected by bacterial virulence factors, in some experiments, we used the avirulent aroA S. typhimurium strain. Interestingly, we found that aroA S. typhimurium strain also induced an expansion of CD8
+TCR
+ iIEL population to a similar extent as wild-type Salmonella (Fig. 2C).
Salmonella-induced expansion of the CD8
+TCR
+ iIEL subset is class I and TAP dependent but MHC class Ia independent
Previous studies have demonstrated that the development and function of TCR
+ iIELs is largely MHC class I and thymus independent. However, CD8
-expressing T cells have been shown to be MHC class I and thymus dependent, and the recognition of MHC class I molecules by TCR
+ iIELs has been described (39, 40, 41, 42). To assess the potential involvement of MHC class I molecules in the expansion of CD8
+TCR
+ iIELs in response to Salmonella infection, MHC class I-deficient (
2m/), TAP-1-deficient (Tap-1/), and MHC class Ia-deficient (Kb/Db/) mice were orally infected with wild-type Salmonella. The expansion of a CD8
+TCR
+ iIEL population of similar size was observed in C57BL6 and Kb/Db/ mice, but not in total class I-deficient
2m/ mice (Fig. 3A). Thus, the expansion of the CD8
+TCR
+ iIEL population is MHC class I dependent but MHC class Ia independent. In addition, the CD8
+TCR
+ subset fails to expand in TAP-1-deficient mice (Fig. 3B). These results implies that MHC class Ib molecules probably are involved in the TAP-dependent expansion of the CD8
+TCR
+ iIEL subset.
|
The murine class Ib gene products T13, T22, and T23 (Qa-1) have been demonstrated to act as ligands for TCR
+ cells (41, 42, 58, 59, 60, 61). To address whether Qa-1 is involved in the Salmonella-dependent expansion of the CD8
+TCR
+ iIEL subset, a MHC class I-deficient-mouse in which Qa-1 expression was selectively restored was generated. This was accomplished by generating a Qa-1 transgene where
2m was covalently attached. This cis expression of
2m would permit only the Qa-1 molecule to be expressed in
2m-deficient cells and would allow for the identification and characterization of Qa-1-dependent immune-mediated events (52, 53).
The scQa-1b transgene was generated by connecting the C terminus of murine
2mb cDNA to the N terminus of the
2 domain of Qa-1b cDNA (scQa-1b(c)) or Qa-1b genomic DNA (scQa-1b(g)), via a spacer of 15 aa, containing four glycines followed by a serine, repeated three times (G4S)3 (Fig. 4A). The correct molecular mass of scQa-1b was verified by Western blot analysis of whole-cell lysates from cells stably transfected with scQa-1b(c) or scQa-1b(g) (Fig. 4B). Results obtained from scQa-1b(c) and scQa-1b(g) were equivalent. No significant proteolytic degradation of scQa-1b was observed, demonstrating that it is efficiently synthesized and transported to the cell surface.
|
2m/ cell lines were performed. As shown in Fig. 4C, scQa-1b(c), when transfected into
2m-deficient cell lines DLD-1 and KJ29, reaches the cell surface at significant levels; studies with scQa-1b(g) showed similar results (data not shown). Furthermore, scQa-1b and wild-type Qa-1b were effectively recognized by the Qa-1-restricted CTL clone 39.1D7X as well as by the Qdm-dependent Qa-1-restricted CTL clone D5D2 (Fig. 4D). Because 39.1D7X recognizes a processed endogenous peptide, these results indicate that scQa-1b can serve as a functional Ag-presenting structure. D5D2 only recognizes Qa-1b on the surface of C1R cells in the presence of exogenous peptide. Therefore, scQa-1b is able to bind exogenous or endogenous peptide for recognition by Qa-1-restricted CTLs with similar efficiency to Qa-1b.
Transgenic mice were established using the two scQa-1b constructs, scQa-1b(c) and scQa-1b(g). One scQa-1b(c) and two scQa-1b(g) founder lines which express scQa-1b were selected for further analysis. ScQa-1b expression was observed in the small intestine, spleen, and thymus of scQa-1b+/
2m/ mice, at lower levels than Qa-1b (see Fig. 7B). The results obtained from all founder lines were equivalent, and the data from one of the scQa-1b(g) founder lines is shown.
|
2m/ mice was analyzed by assaying the ability of lymphoblasts from wild-type (B6), scQa-1b+(g)/
2m/, and
2m/ mice to serve as targets for lysis by the Qa-1-restricted CTL clone D5D2. As shown in Fig. 4E, lymphoblasts from wild-type B6, scQa-1b+(g)/
2m/, but not
2m/ mice, were recognized by D5D2. Hence, scQa-1b is functionally expressed on the surface of lymphoblasts.
Salmonella-induced expansion of CD8
+TCR
+ iIEL is Qa-1b dependent
The Qa-1 dependence of the expansion of the CD8
+TCR
+ iIEL population in response to Salmonella infection was addressed by orally infecting wild-type,
2m/, and scQa-1b+/
2m/ mice with wild-type Salmonella. Flow-cytometric analysis revealed that the CD8
+TCR
+ iIEL population expanded in the scQa-1b+/
2m/ mice in a similar fashion to that observed in wild-type mice (Fig. 5). Therefore, selective restoration of Qa-1b expression in class I-deficient animals restores the Salmonella-driven expansion of the CD8
+TCR
+ iIEL population.
|

+TCR
+ iIEL population (Fig. 6). Hence, the Qa-1-dependent expansion of this population must occur independently of interactions with CD94/NKG2.
|
The Qa-1-dependent expansion of the Salmonella-induced CD8
+TCR
+ iIEL population implies that Qa-1 is expressed in the intestinal epithelial compartment, and thus is available for recognition by other cells. To directly address this prediction, Qa-1 expression in the small intestine of B6 mouse was examined by both Western blot and immunohistochemistry. As shown in Fig. 7A, Qa-1 expression can be readily detected in the spleen and thymus, and high-level expression is also observed in the small and large intestine. In addition, the scQa-1b transgenic protein is also expressed at high levels in the intestine (Fig. 7B). Immunocytochemical analysis detected Qa-1 expression in intestinal epithelial cells lining the villi (Fig. 7C). Qa-1 expression appeared largely intracellular and was absent in goblet cells, and lower expression levels occurred in undifferentiated crypts cells. Thus, Qa-1 is expressed at high levels in intestinal epithelial cells and is available for interaction with 
iIELs.
| Discussion |
|---|
|
|
|---|

+TCR
+ iIEL population is induced. This subset can be detected as early as 3 days following infection and by 1012 days can represent 1015% of the total TCR
+ iIELs. The induction of these CD8
+TCR
+ iIELs seems to be independent of overall bacterial virulence, because oral infection with wild-type and aroA Salmonella strains can induce this subsets appearance. To the best of our knowledge, this report is the first description of a class I-dependent CD8
+TCR
+ iIEL subset, because previous studies have demonstrated that iIELs expressing TCR
are either CD8
or CD8 negative, and are class I independent (38). We propose that this CD8
+TCR
+ iIEL subset is an early-acting component of the overall host immune response that will ultimately lead to bacterial clearance.
A resident class I and TAP-dependent CD8
-expressing V
4 cell population has been identified in the lung, which increases following Ag challenge in a murine model for airway hyperresponsiveness (71). This lung population appears distinct from the intestinal CD8
-expressing TCR
cell population we have described, because this iIEL subset is rare in resident cells and fails to express V
4 (S. Lopez-Briones and M. J. Soloski, unpublished data). Therefore, we predict that the lung- and intestine-localized CD8
-expressing TCR
cells will have distinct recognition properties and possibly effector function. However, the finding that CD8
-expressing TCR
T cells expand in the lung and gut settings following Ag exposure, suggests that the expansion of class I-dependent 
T cells may be a general, hitherto unrecognized, feature of tissue-localized 
T cells following an inflammatory stimuli.
The precise nature of the antigenic stimulus that drives the increase in CD8
+TCR
+ iIELs is not completely defined. However, the expansion of this CD8
+TCR
+ iIEL population appears to be dependent on expression of the class Ib molecule Qa-1. This conclusion is supported by the absence of CD8
+TCR
+ iIELs in class I-deficient mice and their presence in class Ia-deficient mice and in class I-deficient mice where Qa-1 expression has been selectively restored. The most straightforward interpretation of our data is that the CD8
+TCR
+ iIELs are directly recognizing Salmonella-induced changes in Qa-1 expression. Qa-1 is recognized by two types of receptors, either clonally expressed T cell Ag receptors or the CD94/NKG2 receptor expressed on NK cells and some T cell subsets (65). Our analysis shows that the CD8
-expressing TCR
T cells do not express CD94/NKG2 NK receptor family members; therefore, direct recognition by TCR
is most likely. Prior studies have demonstrated that Qa-1 can be recognized by TCR
+ cells. Vidovic et al. (60) demonstrated that a 
T cell hybridoma recognizes a Glu50Tyr50 synthetic copolymer presented by Qa-1. Also, Qa-1 can bind and present peptides derived from bacterial and mammalian heat shock protein (hsp)60 family members, and several examples of hsp-reactive 
T cells have been reported (72, 73, 74). Therefore, it is reasonable to hypothesize that the Qa-1-dependent expansion of the CD8
could be TCR mediated. Furthermore, it is likely that the recognition of Qa-1 by the CD8
-expressing TCR
T cells involves peptide, because the expansion of these cells is TAP dependent. This is consistent with the known function of Qa-1 in the presentation of peptide to antibacterial and alloreactive effector CD8+ T cells (21, 51, 70, 75). Indeed, CD8
-expressing TCR
T cells found in the lung are TAP dependent, providing another example in which TCR
T cell recognition is peptide dependent (71).
Qa-1 is readily available for recognition by TCR
+ iIELs, because it is expressed at high levels by intestinal epithelial cells. It is interesting to note that the Qa-1 staining pattern observed in our studies suggests that Qa-1 is predominately intracellular. We speculate that, during infection, Qa-1 surface expression on intestinal epithelial cells is up-regulated, perhaps due to the increased supply of a relevant Qa-1 binding peptide. This peptide would not be derived from the leader sequence of class Ia molecules because CD8
-expressing TCR
cells are found in H2-K/D-deficient mice that lack a source for these leader peptides (76, 77). In contrast, Qa-1 has been shown to bind and present peptides derived from Salmonella and murine hsp60, and these peptides would be likely candidates (21, 75). Such peptides may also be recognized by the CD8
-expressing TCR
iIELs.
At present, the precise origins of the bacterial-induced CD8
-expressing TCR
iIELs are not clear. One possibility is that these cells represent a bacterial-induced expansion of a rare, normally resident iIEL subset or are a population of recent migrants attracted to the epithelial compartment by the infection. Alternatively, they could represent a novel class Ib-dependent subset of CD8
-expressing TCR
iIELs that have up-regulated CD8
following activation. At present, we cannot distinguish between these possibilities, although the up-regulation of CD8
in CD8
-expressing T cells has not been described previously. Class I-dependent, CD8
-expressing TCR
cells are known to originate in the thymus where they undergo selection (78, 79). Whether this observation applies as well to CD8
-expressing TCR
T cells remains to be determined.
TCR
+ cells have been proposed to serve several roles. These functions include immunoregulation of the adaptive TCR
-mediated immune response, direct pathogen-specific effector function and the maintenance of the integrity of the epithelial barrier (80, 81, 82). Previous studies have shown that, following oral infection with the coccidian protozoan Eimeria, a biphasic increase in TCR
iIELs has been detected (83). Because TCR
-deficient mice do not display increased susceptibility, it appears that the Eimeria-induced TCR
+ iIEL subset does not directly play a role in pathogen clearance (84). However, such mice display increased epithelial damage, indicating a role for TCR
+ cells in either regulating TCR
-driven immunopathology or a direct role in the maintenance of the integrity of the epithelial compartment (84). The latter role is supported by the observation that 
T cells are necessary for the preservation of the intestinal epithelium in response to inflammation and that skin 
T cells release keratinocyte growth factor, a factor that stimulates epithelial cell recovery (85, 86, 87).
Infection of 
T cell-deficient mice with either Listeria monocytogenes or Salmonella also indicated that 
T cells play little role in controlling infection (9, 88, 89). However, in the Listeria infection model, 
T cell-deficient mice display an increased inflammatory response and tissue damage (88, 89). Collectively, such studies argue for a role for 
T in controlling the extent of infection-induced inflammation and/or maintaining tissue integrity (recently reviewed in Refs.80 and 82). After mice are infected orally with Salmonella, the bacteria invade intestine epithelium and then seed the liver and spleen (5). Infection of intestinal epithelium by Salmonella has been shown to induce cell death and thus damage the epithelial barrier (90, 91). Therefore, a role for TCR
iIELs, in particular the CD8
-expressing TCR
cells, in the recovery of the intestinal epithelium from Salmonella infection is reasonable. In contrast, given that these cells express CD8
and are class I dependent, properties typical of conventional TCR
+CD8+ cells such as a direct effector role in limiting bacterial growth and/or clearance cannot be ruled out. In vivo depletion of TCR
+ cells results in an increase in susceptibility following oral infection with S. enteritidis, implying such a role (92).
The class I-like molecules RAE-1, H-60, and MIC-A, and the T10/22 class Ib molecules have been proposed to serve unique roles in the identification of altered cells (93, 94, 95). The expression of these molecules is restricted to certain cell types (e.g., epithelia), and their expression can be induced by stress, inflammation, or infection. Qa-1 may fulfill a similar role, and stress-induced changes in Qa-1 expression have been noted (73, 96). Therefore, Salmonella-induced changes in Qa-1 surface expression on intestinal epithelial cells could be recognized by subsets of TCR
+ iIELs and play a role either in limiting infection and/or in maintaining the epithelial barrier. Considering that Qa-1 functional counterpart HLA-E, as well as other novel class I molecules, are expressed in the gut mucosal environment (97, 98), the recognition and effector properties of this CD8
-expressing TCR
T cell subset will need to be considered as we seek to understand the host-pathogen interaction within the mucosal compartment.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 The authors declare that they have no competing financial interests. ![]()
3 Current address: Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, NY 10021. ![]()
4 A.D. and S.L.-B. contributed equally to this work. ![]()
5 Address correspondence and reprint requests to Dr. Mark J. Soloski, Division of Rheumatology, Ross Research Building, Room 1068, The Johns Hopkins University School of Medicine, 720 Rutland Avenue, Baltimore, MD 21205. E-mail address: mski{at}jhmi.edu ![]()
6 Abbreviations used in this paper: PP, Peyers patch; iIEL, intestinal intraepithelial lymphocyte; TS, tryptic soy; MLN, mesenteric lymph node; scQa-1b, single-chain Qa-1b; hsp, heat shock protein;
2m,
2-microglobulin. ![]()
Received for publication December 17, 2003. Accepted for publication March 18, 2004.
| References |
|---|
|
|
|---|

and 
T cells in the host response to Salmonella infection as demonstrated in T-cell-receptor-deficient mice of defined Ity genotypes. Infect. Immun. 65:2306.[Abstract]
and IFN-
in recall of immunity to oral challenge with virulent salmonellae in mice vaccinated with live attenuated aro-Salmonella vaccines. Microb. Pathog. 13:477.[Medline]

cells and IFN-
in bacterial clearance independent of intracellular location. J. Immunol. 156:3321.[Abstract]
-
T-cell receptor on intestinal CD8+ intraepithelial lymphocytes. Nature 333:855.[Medline]

+ intraepithelial lymphocytes. Science 243:1716.
/
and CD8
/
T cells in major histocompatibility complex class I-deficient mice. J. Exp. Med. 190:881.
+ intestinal intraepithelial T cells in
2-microglobulin- and/or TAP1-deficient mice. J. Immunol. 156:2710.[Abstract]
2-microglobulin. Eur. J. Immunol. 25:2332.[Medline]
/
T cell receptor
/
intestinal intraepithelial lymphocytes in the absence of both classical major histocompatibility complex class I and nonclassical CD1 molecules. J. Exp. Med. 190:885.
+CD8
+ T cells are found in intestinal intraepithelial lymphocytes of mice that lack classical MHC class I molecules. J. Immunol. 163:4100.
/
T cell receptor
/
+ cells in murine intestinal intraepithelial lymphocytes. J. Exp. Med. 192:1521.

T cells develop normally in
2-microglobulin-deficient mice. Proc. Natl. Acad. Sci. USA 89:653.
T cells. Cell 76:29.[Medline]

T cell receptor. Cell 62:549.[Medline]

T cells that recognize an inducible MHC class Ib molecule. Science 287:314.
and the nonclassical MHC class I molecule, TL. Science 294:1936.
cells. [Published erratum appears in 1993 Immunogenetics 38:80.]. Immunogenetics 36:377.[Medline]
2M, MHC class I proteins and CD8+ T cells. Science 248:1227.
-2 microglobulin-H-2Dd molecules positively select CD8 cells of broad repertoire when expressed transgenically in
-2m knockout mice. FASEB J. 12:1779.
2-Microglobulin gene mutations: a study of established colorectal cell lines and fresh tumors. Proc. Natl. Acad. Sci. USA 91:4751.
T cell receptor ligand T22: a truncated MHC-like fold. Science 287:310.
(
) T cells: non-classical ligands for non-classical cells. Curr. Biol. 10:R282.[Medline]

T-cell hybridoma. Nature 340:646.[Medline]

CTL by TL antigen expressed in the thymus. J. Exp. Med. 184:2175.
/
T cells by the class Ib gene product, Qa-1b. J. Exp. Med. 188:961.
cells. Immunology 94:64.[Medline]
8 TCR: a role of the Qa-1 molecule. Immunity 2:185.[Medline]
4+ pulmonary T cells regulate 
T cell-independent airway responsiveness. Proc. Natl. Acad. Sci. USA 99:8850.
T lymphocytes. Science 249:67.
cells: a large, diversified T-lymphocyte subset with highly focused specificity. Proc. Natl. Acad. Sci. USA 89:4348.
T cells. Nat. Rev. Immunol. 3:233.[Medline]

T cells: functional plasticity and heterogeneity. Nat. Rev. Immunol. 2:336.[Medline]

+ and 
+-deficient mice display abnormal but distinct phenotypes toward a natural, widespread infection of the intestinal epithelium. Proc. Natl. Acad. Sci. USA 93:11774.
T cells in tissue repair. Immunol. Res. 21:63.[Medline]

T cells. Proc. Natl. Acad. Sci. USA 99:14338.
T cells in wound repair. Science 296:747.
and 
T cells in immunity against an intracellular bacterial pathogen. Nature 365:53.[Medline]

T cells. J. Immunol. 153:3101.[Abstract]

T-cell antigen receptor contribute to resistance to Salmonella infection in vivo. Infect. Immun. 62:4618.
T cells. Science 294:605.
T cells. Science 279:1737.This article has been cited by other articles:
![]() |
H. Chen, X. He, Z. Wang, D. Wu, H. Zhang, C. Xu, H. He, L. Cui, D. Ba, and W. He Identification of Human T Cell Receptor {gamma}{delta}-recognized Epitopes/Proteins via CDR3{delta} Peptide-based Immunobiochemical Strategy J. Biol. Chem., May 2, 2008; 283(18): 12528 - 12537. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Chattopadhyay, J. O'Rourke, and R. E. Cone Implication for the CD94/NKG2A-Qa-1 system in the generation and function of ocular-induced splenic CD8+ regulatory T cells Int. Immunol., April 1, 2008; 20(4): 509 - 516. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. B. Ascon, S. Lopez-Briones, M. Liu, M. Ascon, V. Savransky, R. B. Colvin, M. J. Soloski, and H. Rabb Phenotypic and functional characterization of kidney-infiltrating lymphocytes in renal ischemia reperfusion injury. J. Immunol., September 1, 2006; 177(5): 3380 - 3387. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. A. Ingersoll and A. Zychlinsky ShiA Abrogates the Innate T-Cell Response to Shigella flexneri Infection Infect. Immun., April 1, 2006; 74(4): 2317 - 2327. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. A. Guidry and I. Stroynowski The Murine Family of Gut-Restricted Class Ib MHC Includes Alternatively Spliced Isoforms of the Proposed HLA-G Homolog, "Blastocyst MHC" J. Immunol., October 15, 2005; 175(8): 5248 - 5259. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |